How Generative AI is Affecting Fractal Art
Generative AI is reshaping the role of fractal artists and the art they make, but is this the next step
in the evolution of this art form or, as a lot of people see it, are we heading down the wrong path?
Artist communities on social media are divided between those who distinguish between AI-generated
"fractal-like" images created using prompts and those created through traditional techniques involving
the use of software to explore, frame, and add definition to fractals generated by recursive
mathematical formulas.
Many communities either embrace AI art, require it to be labelled as such, or specifically exclude it
altogether.
Artists who oppose AI argue that while AI can produce stunning, complex, and high-quality images that
mimic intricate fractal patterns, these images are not 'true' fractals.
Their reasoning is that AI lacks an understanding of the underlying mathematics, resulting in images that
are representations of fractals rather than actual mathematically precise self-similar objects.
But are the fractal "purists" correct? After all, if an artist's work possesses fractal-like qualities,
such as self-similarity at different scales or infinite detail, shouldn't it be considered fractal art,
regardless of how it was created?
Strictly speaking, AI-generated fractal art aligns with the concepts outlined in the Fractal Manifesto
published in 1999 by Kerry Mitchell. This manifesto has served us well for over 30 years and defines
fractal art as a distinct, legitimate two-dimensional art form, often created using mathematical
algorithms to produce complex, self-similar, and often digital imagery.
The reality is that if we want to categorise fractal art into two groups—one that includes AI-generated
fractal art and one that does not—we must acknowledge that the boundaries were blurred long ago, even
before the advent of AI.
In fact, the only form of fractal art that could be deemed not to include AI generative AI or be 'pure'
probably only really existed in the original definition from the late 1980s, where Benoit Mandelbrot
described it as "art for the sake of science."
This original concept leaves no room for doubt that the moment fractal art is made without it being the
result of a scientific endeavour, then it stops being fractal.
Art for the Sake of Science
To understand the concept of "Art for the Sake of Science" and its implications for AI-generated fractal
art, we must go back to the late 1980s when fractals first began to be recognised as an art form.
To defend this new medium against scepticism and highlight it as a legitimate form of modern, digital,
and visual artistry, fractal art had to separate itself from other emerging digital art forms.
It gained acceptance as a style of art due to its unique function in explaining and advancing scientific
knowledge. The role of the artist became one of using creative skills to help scientists visualise their
research differently. In the case of fractals, this meant making complex data more accessible and
provoking new ways of thinking.
In his introduction to the fractal art displayed at the SIGGRAPH 1989 art show, one presenter stated,
"Can a shape that is defined by a simple equation or a simple rule of construction be perceived by
people other than geometers as having aesthetic value - namely, as being at least surprisingly
decorative - or perhaps even as being a work of art? When the geometric shape is a fractal, the answer
is yes."
In many ways, fractal art has evolved beyond this original idea, and an app that generates a fractal is
doing no more or less than AI if its sole purpose is to create a fractal image.
Defining fractal art too narrowly could limit the potential of AI-powered tools that have the capability
to push the boundaries of this medium.
Furthermore, making such distinctions raises complex questions. For instance, what happens if you use AI
to operate a fractal-generating tool? Is that considered permissible in traditional fractal art? What
does it mean for those who use AI to write the code that creates their fractal art? And, what if IA is
used by scientists to make art of their scientific contributions?
Ultimately, I would discourage 'Fractal purism', especially since art is always evolving and AI is still
relatively new. However, this does come with a warning.
The Problem With IA-Generated Fractal Art
There is a problem with AI-generated fractal art that parallels challenges faced by the broader arts
community. As AI gets better at making fractals, we unfortunately seem to be getting worse.
The study of fractal geometry combines art and science, offering a unique way to describe complex,
self-replicating patterns found in nature. Fractal images possess both scientific applications and
aesthetic appeal, and the visualisation of fractals has long fascinated both artists and scientists, who
have shared the creation of these images for the benefit of both communities.
Unfortunately, the rise of art generated on AI platforms is overshadowing manually created fractal art
and while AI has made it easier to create fractal artwork, it also omits content that highlights the
significance of fractals, moving this understanding further beyond the reach of most artists.
As AI-generated fractal art is promoted by the algorithms of social media platforms, artists and
scientists who once shared an understanding are being separated.
This AI bias is leading to a loss of understanding among artists about how this art form is
intrinsically linked to scientific discovery.
If we use AI to make fractal art easier, but not to understand it better, a connection between the arts
and sciences will be lost, and it will become meaningless.
If fractal artists become isolated from scientists, this will limit the rewards gained through
collaboration.
And if artists and scientists no longer share the same understanding, they cannot help each other share
their ideas.
